California-based open-source Generative AI company H20.ai has developed a new model risk management framework for generative artificial intelligence, specifically designed for banks and financial services firms.
The firm claims the platform integrates automated testing, human calibration, and bias detection, and a host of big banks have already signed up for the solution, including Wells Fargo, USAA, UBS, HSBC and Barclays.
H2O.ai, which is generally considered to be one of the biggest players in open-source Generative AI and Predictive AI platforms, said the new solution is aimed at improving transparency, compliance, and trust within banks and financial services firms as AI continues to rapidly gain traction in the finance sector.
The firm was keen to stress it has recently signed up major institutions like Wells Fargo, USAA, HSBC, UBS, and Barclays in building AI-driven solutions.
“This new framework provides a structured evaluation methodology that integrates automated testing, human calibration, and bias detection, offering model validation before deployment,” explained Sri Ambati, the chief executive officer and founder of H2O.ai.
“It aims to mitigate risks such as biased decision-making and security vulnerabilities, crucial concerns for financial services operating under strict regulatory guidelines,” he added.
‘Trustworthy’ AI for banks
Ambati said he spotted an appetite for his new product in the market. “With the growing adoption of Generative AI in sensitive areas like banking, it’s more important than ever for financial institutions to ensure that AI models are both reliable and transparent,” he said.
Ambati stressed H2O.ai’s framework “brings this to life” by enhancing existing risk management practices and extending them to the complexities of Generative AI.
Banks will be offered access to a full suite of tools to validate AI models, including automated test generation.
“It generates diverse test cases using advanced modeling techniques, ensuring broad coverage of potential inputs,” he noted.
“Regulated industries need trustworthy AI that meets strict compliance, risk, and transparency requirements.”
– Sri Ambati
Then there is human-calibrated evaluations. “It aligns machine evaluations with human judgment to ensure real-world applicability,” Ambati said.
Moreover, the model identifies model weaknesses and implements safeguards to avoid failures or security breaches.
Finally, there is robustness testing as the simulation simulates real-world conditions by testing AI models against adversarial inputs and out-of-distribution queries, Ambati shared.
Regulatory compliance
Ambati continued by saying that can confidently deploy AI solutions with auditable and reliable results, as long as they run sufficient tests.
“The integration of the latest MRM tools will enable leading financial services companies such as Comerica, Northern Trust, and HSBC to comply with stringent regulations while minimizing third-party validation risks,” he said.
“Regulated industries need trustworthy AI that meets strict compliance, risk, and transparency requirements,” said Ambati.
He added that the company has trained risk teams and model validators at top banks like CBA, KeyBank, and US Bank, building in-house AI expertise that reduces reliance on external third-party vendors.
“This enables quicker, safer, and more cost-effective deployment of AI models across the financial sector,” he concluded.
Why not become a QA Financial subscriber?
It’s entirely FREE
* Receive our weekly newsletter every Wednesday * Get priority invitations to our Forum events *
THIS MONTH
QA Financial is delighted to announce that Tal Barmeir will join us as a speaker at the QA Financial Forum Toronto 2025 Places are limited – register today.
WATCH NOW
READ MORE
- Taking place tomorrow: the QA Financial Forum Chicago 2025
- HDFC Bank turns to Katalon and QualityKiosk for QA upgrade
- Perforce’s Clinton Sprauve on AI testing for charts and graphics
- Functional and regression testing crucial for banks, says insider
- Shastic teams up with MeridianLink to target banks with AI automation
